U.S. Digital Service recruits Silicon Valley innovators, like Matt Cutts, to modernize government technology

I recently interviewed Matt Cutts, who is the Acting Administrator at US Digital Service, a non-partisan technology group in the Executive Office of the President whose ambitious goal is to deliver better government services and improve the lives of all Americans. Matt initially joined US Digital Service in 2016 at the Department of Defense. Since 2017, he has served as Acting Administrator and is responsible for setting the overall direction and strategy for agency projects. Previously, Matt was an early employee at Google (joining in 2000). He was head of the webspam team, where he protected the quality of Google’s search results and answered questions about search engine optimization and ranking algorithms. He also wrote the first version of SafeSearch, Google’s family filter. Matt is widely known for his 30-day challenges to try something new and is an avid (but slow) long-distance runner. Matt holds a master’s degree from the University of North Carolina at Chapel Hill, and bachelor’s degrees in both mathematics and computer science from the University of Kentucky.

US Digital Service

So – what is United States Digital Service?

Matt: Sure, that’s a great question! Our most well-known piece of reputation is when HealthCare.gov fell over and needed to be revived, and so a team of geeks came in and helped get that system back up again. And so, the US Digital Service is a group of designers, engineers, and product managers who bring the best practices from industry into the government to try to make services work better for the American people.

Can you give us an example or a use case that’s public that we could grasp onto to understand more?

Matt: Sure, totally. So, an example that I love to talk about is like a veteran is trying to access for health benefits. They needed to fill out a form called the 10-10EZ, and unfortunately, it wasn’t very easy. You had to have the right version of Internet Explorer, and it had to be Internet Explorer. It couldn’t be another browser like Safari or Firefox or Edge, and you had to have the right version of Adobe Acrobat.

And if you didn’t have the right version of both of those, you would get an error message that says you need to upgrade your version of Adobe Acrobat. And that was actually wrong; you needed to downgrade your version of Adobe Acrobat. So only about 8% of veterans were able to apply for the health benefits online, and when they used the paper version of the form, it takes on average 137 days for them to get their benefits.

So, we implemented a simple web form. It’s responsive, so it works on mobile phones. It’s accessible, so screen readers and blind users can use it. And 50% of veterans find out whether they’re eligible for benefits within ten minutes. So, we’ve now had 350,000 plus veterans apply for their health benefits using this form. When you explain to technologists and designers that incredibly straightforward, simple changes can have a radically huge impact on someone’s life they really get excited and interested in it.

That’s so incredible!  So, let’s talk about how you pick projects.  18F, for example, has a traditional sort of consulting model, whereas United States Digital Service has a … is funded by almost like a block grant. So, your shop is able to pick projects, whereas these folks are sort of searching for work.  What is your criteria for picking a project, and what sparks your interest?

Matt: Totally. We have three basic criteria when we look at projects to decide whether we can engage. The first one is the impact. What is the overall difference that it’s going to make to the American people? And how visceral is that impact?

The second one is, what does the executive support look like? Because it really helps if there’s an escalation path or a sponsor or someone who cares deeply about the issue of who can help to clear the way and move obstacles out the way.

And then the third thing that we look for is how reusable is that solution? Can that solution be used in other parts of the government, or can it scale up in order to have a large impact. So, if we see a project hit on all of those conventions in a good way, then we start to get excited about how we can potentially work on that project.

That’s awesome! Really, really neat. And so, beyond the veteran’s project, give me an example of a project that fell in line with those three things.

Matt: Sure. So, HealthCare.gov was interesting because it was a crisis.

But as a result of HealthCare.gov, the folks that we worked with at health and human services and CMS recently had a law that was passed that would have affected them, it was called MACRA. And the idea behind MACRA is to move from fee for service. So, the number of tests that somebody runs, and how expensive those tests are, and being reimbursed for that, to value-based care.

So, outcomes like—if you are discharged from the hospital, are you readmitted to the hospital within thirty days. And this was a law that had a hard deadline, January first of this year was the beginning of the submission period for positions. And so, CMS came to us over a year ago and said, ‘How can we figure out how to do this better?’

And so, if you look at the playbook of what United States Digital Service does well, we helped not only designing the system, we also helped on writing the requirements. So, CMS typically would take a year to write requirements, go through a public comment period, and then finalize the regulations.

And what we did working with CMS is to say, ‘why don’t you release an initial draft set of comments in one-fifth of the time?’ Get responses back, we can mock up websites to do those sorts of, those policies, and get reactions. And then we could iterate, and we could do another take at doing the regulations.

And what happened is, after a year they were able to do five iterations, and it turns out if you get five practice shots at a goal you’re probably going to be more likely to hit it when you take your real shot. So, we talked to people at CMS who said, ‘this was fantastic, we’re never going to do it the old way again, we really think this ends up with better regulations and better results.’

We also helped find qualified contractors. So, a lot of contractors are typically rewarded for how well they can write their proposals, and what we decided to do is ask contractors to write code and then we used engineers at the United States Digital Service to simply evaluate the quality of that code. 

And so, we ended up with a contracting vehicle and agile purchase, blanket purchase agreement, a qualified pool of contractors, better regulations … and then we also helped to coordinate all of that along with our partners at CMS.

So, the quality payment program, as it’s known, launched January first of this year. The submission period just recently ended, and the beauty, the wonderful thing about all of it is that we didn’t hear a single thing about this program because it didn’t have any hiccups.

We probably should give awards for when nothing happens, because the project was under budget, it saved money, it’s better for its positions, they’re more likely to stay enrolled in Medicare. And it’s just the kind of effective and efficient government that you’d like to see more of.

Simply amazing!  So, let’s switch gears.  During the US House Committee session on Blockchain technology …  it was interesting because Congresswoman Bonamici said she had gone to Estonia and looked at blockchain for e-voting. Congresswoman Esty referenced Crypto Valley in Switzerland and China Open Trust Blockchain Communities; then I was with the head of UK Government Digital Services Kevin Cunnington, and he said you guys (GDS and USDS) have a global network of folks who are working together. So, could you talk a little bit about your engagement in the global network and maybe two or three learnings received from the global network?

Matt: Absolutely. So, we’re currently at the Code for America summit and I was just in a panel from, that involved representatives from lots of other digital services. So, GDS, the Government Digital Service in the UK. CDS, the new Canadian Digital Service. And it’s interesting to be able to swap ideas and techniques, and even bureaucracy hacking, which is to try to make things work better.

So, for example, we were just recently talking about how bringing users into the conversation, even as simple as recording a user trying to use a consumer service can often make good progress on trying to deliver support behind reforming things or making a change. And a person from another digital service made a very interesting point, which is 70% of the time, it might work great to say, ‘think about the users, and what their experience is like.’ But maybe 30% of the time, people think about costs and economics.

And so being able to make all the different arguments. The fact that modern government IT can be not only better, but also cheaper, and it can be delivered faster. They’re … it’s like flip sides of the same coin but figuring out what argument resonates with somebody is often a great way to figure out how to make forward progress.  So that’s just a very simple example where trading ideas and being able to participate in that global dialogue can really help.

US Digital Service

That is awesome because, during this 4^th Industrial Revolution, the world is facing massive problems, due to a lack of technical talent, which lead to the rise of digital services teams.  How does United States Digital Service fit into the workforce of the future?

Matt: One thing that I think is very interesting is the idea of re-skilling and retraining people. So, if we can free up people to work, not on the tedious or the mundane things, that allows them to then spend their time on the more interesting, or the more critical cases.

So, for example, we were recently working with green cards. And, it turns out that there was a pretty long backlog for approving the renewal of green cards. And after doing some discovery work, what we found is that about half of all of the people in the backlog didn’t need to go through any extra checks. The name was the same; the address was the same. Nothing had really changed for them, there were no security issues.

And so rather than having an adjudicator look at that in person, by hand, you could auto-approve those. Or maybe approve them in a batch, with several different ones. And what’s great about that is that then makes the workload easier for the adjudicators, the smaller cases that they don’t have to spend as much time on. They can take those off their plate, and work more on the more interesting, more important, more critical cases.

So, I think it’s always important to be mindful of the impact of technology on people. And I think that governments across the world are trying all kinds of experiments, and the beauty is that these sorts of laboratories of different approaches also act as existence proofs. So, if you see something working really well in Estonia, that shows that it’s possible. A friend from the Canadian Digital Service said that could be a landmark. It doesn’t give you the roadmap to have where you want to go, but you know that it is possible to get there.

And so, just having some of those existence proofs is really useful.

Dr. Emma Chory and I always debate why would someone let go of the incentives and salaries large tech companies provide to study simple sciences. So, what would be your call to action to the guy at Google, to the guy at IBM, to the guy at Facebook to join the US Digital Service?

Matt: Absolutely. First and foremost, I think it would be about mission and purpose. A lot of people are looking for meaning right now. If you’re sitting at a desk at Google, you’re one 70,000 people. If you want to come to US Digital Service, the desk might not be as nice, the lunch might not be free, but you’re one of 175 people. And so, the impact that you can have is outsized. 

I told a couple of stories, but we have dozens of stories on our website at usds.gov. We have a report to Congress that details all the things that we’ve been able to work with agency partners to launch in 2017.

So, first and foremost I would say, if you take a look back in the last year, the technology industry has had things like the #MeToo movement. Areas where the technology industry does not necessarily do well, sometimes government does very well. We actually have an incredibly diverse team. 60% of our leadership positions are women. We want to reflect the diversity of America, and we think we do better on our mission when we do reflect the diversity of America.

So, there are amazing things that people can do if they want to come to do a short tour of duty at the US Digital Service.

It’s interesting because a lot of people who might think about doing a tour of duty might feel like an imposter, or ask themselves ‘what do I have to contribute to my government?’

The fact of the matter is if you can code, you can design, if you’re a product manager, there are amazing areas where you can have an impact. And so, a lot of people might not immediately see the path towards where they can join the government. But it’s not as risky as people think.

A lot of companies have civic service leave programs. You can take six months off. Microsoft lets you take up to 13 months off. So, it’s not risky. It’s not a big jump. It’s actually more of an adventure that you can go to learn in DC.

And what I have discovered is that people tend to grow enormously. Like a graduate level course in interpersonal relations in some sense. And yet at the same time, a ton of people have changed their lives. They think that they’ll go back to Twitter in six months, and they discover they actually want to stick around and help make government work better for the people.